Economist cautions AI job-loss studies may overstate harm due to occupation vs. person mismatch
paulnovosad · x · 2026-09-16
Economist Paul Novosad, discussing the Brynjolfsson et al. 2026 paper on AI exposure and employment, urges distinguishing occupations from people:
- If Johnny is hired as Account Manager instead of Data Analyst, statistics show "catastrophic job losses in AI-exposed sectors" while Johnny is fine.
- Young people entering new jobs is the best mode of economic change.
- He notes this is a widespread issue in diff-in-diff literature—implicit assumptions about which group moves normally—the "China shock" papers share the flaw.
Related event: Economists Debate Framing of GenAI Wage Studies(10 posts)→
More from AGI Musings
- Economist proposes using LLMs as 'lab rats' for human behavior research — austinc3301 · 2026-09-16
- Matthew Green: Local Models Will Approach Today's Frontier Capability in a Few Years — matthew_d_green · 2026-09-16
- Demis Hassabis receives RSA Albert Medal, says arts will shape the AGI era — demishassabis · 2026-09-16
- Why the Likeliest AI Risk Scenario Is Not Skynet — 3Quondam6extanT9 · 2026-09-16
- Michael Nielsen Resurfaces Cliff Stoll's 1995 'Internet? Bah' Blunder as an AI-Skepticism Cautionary Tale — michael_nielsen · 2026-09-16
- 60% of Trump voters, 70% of Harris voters see moderate AI risk — Polymarket · 2026-09-16